Yeah this seems like one of the easiest access models to create an excellent security model for -- basically just static publishing. Sounds like they were in need of a rewrite anyway!

It probably seems daunting but to be honest this feels like a weekend's work at this point with LLM assistance. Not to be glib!

▲

jaredwiener 5 hours ago | parent [-]

But it also destroys the business model behind the site.

Sure, you could technically redesign to handle the bot traffic, but if the bot traffic is just taking the data and reducing any need for humans to visit the site, why is he putting in the effort to maintain the site?

▲

tptacek 4 hours ago | parent | next [-]

The business model of the site is apparently private data sales, not ad revenue.

▲

jaredwiener 3 hours ago | parent [-]

Bots don't make purchasing decisions.

▲

Terr_ 2 hours ago | parent [-]

Some people are already trying to make that happen, but I'm convinced it'll be a net-negative.

	▲	jaredwiener 2 hours ago \| parent [-]
		Rephrase: Bots SHOULDNT make purchasing decisions.

▲

hyperhello 5 hours ago | parent | prev [-]

Well, why is he? If it's for ad revenue, this problem is surely global to the web. If it's because he likes to do it, it shouldn't matter if one AI or a trillion scrape the site. I feel like we badly need a rethink of the web architecture after DoubleClick anyway. Maybe the name of the game should be to cut down a site's assets very hard and use static hosting for them. This interferes with crummy sites that show a mess of inlined ads every refresh, but now that there are a billion 'poor users' this is no longer feasible.

	▲	jaredwiener 5 hours ago \| parent [-]
		Maybe more an inherent problem with these chatbots. LLMs are great, but they aren't producing new information. You still need people for that. But if you cut down any incentive for the people to do that, the LLMs will starve.